Replacing fat with carbohydrates decreases both low-density (LDL) and high-density lipoprotein (HDL)-cholesterol. It has been argued that some of the benefits of reducing LDL by lowering fat intake are offset by decreases in HDL. Replacing saturated with unsaturated fat instead of carbohydrates may reduce LDL-cholesterol without reducing HDL- cholesterol, and thus promote more favorable lipoprotein changes. We propose to study the effects of altering dietary fat intake on five HDL subclasses from data on 132 men who spent three weeks on a high-fat reference diet followed by a three-week low-fat reference diet. They were then assigned at random to one of three four-week experimental diets: a monounsaturated enriched diet (N=45); a carbohydrate-enriched diet (N=43); or a high-fat reference diet (N=44). Dietary information from four-day records showed good compliance to the nutritional goals within each experimental diet. Images obtained from nondenaturing polyacrylamide gradient gel electrophoresis will be used to separate HDL into HDL3c (7.2-7.8nm), HDL3b (7.8-8.2nm), HDL3a (8.2-8.8 nm), HDL2a (8.8-9.7nm) and HDL2b (9.7-12nm) subclasses. Specifically, data collected by Dr. Ronald M. Krauss as part of his recently completed randomized controlled trial has been provided to Dr. Williams for the following objectives: Study Aim 1: Confirm that replacing dietary fat with carbohydrates decreases HDL2b, HDL2a and HDL3a and increases in HDL3b. Study Aim 2: Compare changes in HDL subclasses when altering fat intake from 40% to 20% with changes observed when altering intake from 20% to 30% fat. Study Aim 3: Test whether replacing saturated and polyunsaturated fat with monounsaturated fat instead of carbohydrates yields beneficial reductions in LDL-cholesterol without significantly reducing HDL2b or increasing HDL3b. Each study aim will be examined in the total sample and in two genetic polymorphisms that have been shown to affect lipoprotein responses to diet: 1) LDL-subclass pattern and 2) apolipoprotein E isoforms. The proposed analyses extends our previous studies on diet and HDL- subclasses.